How we read the evidence
Most breathing studies are small, uncontrolled and short. What each sentence in the app rests on, and what we deliberately don't write.
6 min read
Most of the sentences you read online about breathing exercises claim more than the study behind them says. In the app and on this site we try not to do that. Here is how we decide.
Three separate levels of confidence
The things that can be said about a technique aren't equally solid. We divide them roughly into three.
A measured physiological effect. Heart rate variability increasing at close to six breaths a minute, for instance. Repeated, understandable in how it works, measured with instruments. We write this directly.
A result seen in a controlled study. The physiological sigh coming out ahead of other methods in mood measures in a one-month study, for instance. There is a comparison group and a defined dose. We write this naming the study: "in this study, under these conditions, this was seen."
What tradition says. The cooling breath cooling the body, Kapalabhati cleansing, and so on. We either don't write these at all, or write them while saying they belong to tradition.
The common problems of breathing research
Small participant numbers. Studies of twenty or thirty people are common. At that size a chance result comes easily.
Weak control groups. A comparison of "a group doing breathing exercises against a group doing nothing" may be measuring the effect of doing something, not of the breathing. Good studies use an active comparison group.
It can't be blinded. In drug trials the person doesn't know which group they are in. In a breathing exercise they do. Separating out the expectation effect is very hard for that reason.
Self-reported measures. "I felt calmer" is valuable data but it isn't a number measured with an instrument.
Undefined dose. Many studies don't state clearly how many minutes, how many days, or at what pace. A study that doesn't say can't be turned into an app.
What that means in the app
Cooling Breath. Tradition calls it "cooling". In Telles and colleagues' 2020 study body temperature was measured and did not drop. The app says so plainly; the store description never claims "it cools the body" either, only describing the sensation of coolness in the mouth — that sensation is real, the cooling isn't the thing that was measured.
Left and Right Nostril Breathing. The source studies measure single-nostril practice and their participant numbers are small. The text in the app says this openly, along with the weakness of the evidence and the fact that the second values aren't found in the sources.
Right Nostril and the night. There are findings that right nostril breathing increases alertness, but they are weak. So the technique isn't removed from the suggestions at night; only a soft note is added. In the Energizing Breath the note is firmer, because the evidence there is stronger.
Dose. In Left and Right Nostril, the suggested length comes not from the person's answers but from the length applied in the studies. The reason: in these techniques, going below the dose isn't "doing a bit less", it is not doing the thing that was measured.
Sentences we don't write
Sentences of this kind are deliberately absent from the app and the site: "cures anxiety", "lowers blood pressure", "strengthens your immune system", "treats insomnia", "removes toxins".
Some of them may be true. But they aren't things the evidence in hand lets us say, and it isn't right for an app to make medical promises to its users.
Soluklan is a relaxation and focus exercise app. It doesn't diagnose, it doesn't treat, and it doesn't take the place of a doctor's advice.
Where we get it wrong
This approach isn't flawless. When we couldn't find a technique's threshold values in the sources, we have taken them over from similar techniques. Some numbers — how many seconds in, how many out — come from traditional teaching and have no measured counterpart.
In places like that the app's own text says so. When we don't know something, the sentence we write is "we don't know".
Source links in the articles
Every sentence resting on a study has that study's address under it. Most go to a PubMed record or the publisher's full-text page — you can look for yourself and compare it with what we wrote.
Where we couldn't verify a source's full record, we put a search link instead of a direct one. We would rather send you to a search than to the wrong study.